Changelog
Continuous-improvement log of new sectors, signal types, dashboards, methodology refinements, and API additions. Newest first.
Published a full breakdown of the 4-tier confidence model (Platinum/Gold/Silver/Bronze), 7 signal classes (exclusion linkage, enforcement adjacency, registered-agent webs, address clustering, officer overlap, temporal anomalies, asset signals), and the 4-level evidence framework (source-cited, resolution-graded, analyst-reviewed, lead-vs-case). Plus an explicit limits section — what Sentinel does not and will not claim.
All 8 fraud sector dashboards (healthcare, disaster, workers' comp, housing, SNAP/EBT, SBA, unemployment, cross-sector rings) now ship with state-level log-scale heat maps, top-25-by-fraud-amount panels, Sankey-style cross-sector flow visualization (242K cross-pairings tracked), and stacked-bar case-filing time-series 2015-2026.
Visual Explorer graph now includes named individuals connected to firms via OWNS_OR_CONTROLS (Member/Owner/Partner/Managing Member) and AUTHORIZED_OFFICIAL_OF (President/VP/Director/Officer) edges. Enables LittleSis-style power-mapping for fraud-cluster investigation.
Expanded the relationship graph with normalized address clustering across all four customer-lead tiers. Mail-drop and CMRA addresses are flagged separately. Used in combination with other signals to surface shell-network patterns.
Customer leads now resolve to interactive graph views showing the connected entity network. SVG-rendered for fast load, no Cytoscape dependency. 11 node types and 11 edge types color-coded; tier coloring matches in-app lead-tier system.
Healthcare, disaster relief, workers' comp, housing/grant fraud, SNAP/EBT, SBA, unemployment, plus a dedicated cross-sector rings dashboard. ~8.9M entities tracked, $1.13T cumulative exposure across programs.
All customer-facing surfaces now classify entities as either `new_lead` (unprosecuted patterns surfaced from public records) or `proven_case` (already-adjudicated subjects in our reference catalog). The two are surfaced separately and never conflated. Curated case-study materials draw exclusively from proven-case set.
242K entities now indexed as appearing in multiple fraud sectors simultaneously. Highest-signal cross-pairings include healthcare ↔ workers' comp (provider mills), disaster ↔ SBA (PPP/EIDL contractor fraud), and SNAP ↔ housing (multi-program eligibility schemes).
Fortune 500 firms, federal contractors, and large public-sector entities are now systematically suppressed from customer-lead queues. Sentinel targets local fraud — small entities, regional networks, individual operators — not cleanup-penalty matters at major corporations. Blocklist is updated continuously as patterns surface.
REST endpoints for lead detail, sector aggregates, graph neighborhoods, and watermarked dossier delivery. Per-lead model — no bulk export at any tier, by design. See API docs for shape and rate-limit calibration.
Active customers receive briefings on roadmap priorities and beta signals. We don't publish detailed forward-looking commitments here — but we'll tell you what we're working on if you ask.
Request a BriefingImportant disclaimer: SentinelGraph provides structured intelligence based on publicly available data. It does not provide legal advice, conduct investigations, or make determinations of fraud, liability, or regulatory violation. All findings are preliminary, evidence-traceable observations intended to support — not replace — qualified legal, investigative, or compliance review. Use of this service does not create an attorney-client relationship or any professional engagement unless separately agreed in writing.